Identification of models for robust control design

There is a growing interest in the use of system identification as a means to build models for control design. The traditional identification methods deliver a model in the form of a linear time-invariant finite dimensional system. Such a model, called a nominal model, is bound to only approximately describe the dynamics of the plant of concern. In order to cope with this approximation we have to call on methods for robust model-based control design, in order to reach the goal of high performance plant control. It has been widely recognized that not every controller suited for the nominal model will perform equally well with the plant. The general principle is that robustness must be traded off versus nominal performance. We take as a starting point that the problem of constructing a high performance controller for a plant with unknown dynamics, is going to be tackled through identification and model-based control design